Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I don't really understand why Patreon needs so many employees. We're talking about almost 300 people to just supervise a fully automated process. The tech has already been built almost a decade ago and hasn't changed much, so it can't require that much engineering work. It's also an easy problem (take X amount of money, divide it between Y people, pay them out) and they ironed out any potential issues over the years.

For what it's worth I've stopped using Patreon when they included some Facebook tracking (among others) on the membership management page that when blocked will crash the entire front-end and make the entire page unusable.



I feel like this sort of comment can only come from somebody who's never had to deal with operations.

When you have a lot of users (especially when they rely on you to get money from you!) even 0.1% of them asking questions and having difficulties is a huge work generator!

It's not "oh it's easy to write a for loop". It's "my for loop now has 20 non-orthogonal options, and a user is saying we emptied their bank account and now we gotta make sure if we did". Also a content host, a CMS, and reporting software for taxes and the like.

I mean I bet you could get away with 100 people or something for this but there's _so much_ that can go wrong and would require lots of intervention.


Exactly this.

A lot of people don't seem to realize, when you're dealing with communities -- especially communities that involve money flying around -- there's tons of work in managing it.

You've got employees trying to grow the community in different areas (genres?) and different geographies, reaching out to people, giving advice to get them on board. You've got employees giving support. You've got employees looking for irregularities and fraud among all the finances.

Sales, marketing and operation are huge parts of businesses that many engineers don't often think about.


That argument makes sense when you are Reddit which has 400 employees. Patreon has no excuse.


> I feel like this sort of comment can only come from somebody who's never had to deal with operations.

If you really think that way then the company is doomed. The primary reason for technology companies to exist is the fact that the number of users is disconnected from the number of employees in the company. Companies like Patreon completely waste this huge advantage and just set money on fire. You are no longer a tech company anymore and if you are not a tech then you don't deserve unicorn valuations. This is why the company is doomed.


Yeah it still sounds like a pretty easy problem (I work at a super operationally heavy company and at time my workload has been 90% ops 10% eng). If the 300 count includes all the support agents then I maybe I can see it but the common Silicon Valley practice would be for none of those people to actually be employees of the company.

On the other hand my company has zero marketing, which obviously won't be the case for Patreon - I have no idea what headcount levels are involved there - at Google marketing had a ton of people like running promotions and campaigns but again the entire bottom layer was TVC's...


This is hacker news. Under each article you see people being surprised why companies have employees. Common knowledge is that all you need is to write some code and you can start earning billions!


Or done financial audits. Or moved money, let alone between countries. Or done taxes in multiple countries.


If 0.1% of your customers cause 80% of your effort maybe rethink that retention?


These customers are usually top billing and are just indicators of going upmarket. They’re also probably hitting problems loads of your users are hitting and simply dropping out of the funnel from.

“Fire your users” is a thing, but at one point you gotta actually have things work


Depending on the business, this can be like poking out your eyes to save money on eyeglasses.


That's usually balanced against how much money they bring in. If they are profitable customers then it makes sense to hire enough to support them.


> I don't know why X needs N people

This statement can be made about almost any company in the US right now. There is some serious systemic risk in the surreal number of people working whose jobs do not matter at all.

Even for productive software engineers, think back on how many projects in your career really shipped, how many of those then made any money, and then how many of those companies are still thriving. For me nearly all of the work in my career has amounted to surprisingly little and I have always been on the front lines of creating products in my roles.

So why are there so many people at every company that don't really create value? What does it mean when an economy depends on those people being employeed, well paid and consuming goods?


> There is some serious systemic risk in the surreal number of people working whose jobs do not matter at all.

There is perhaps greater systematic risk in not having backups. Too much efficiency can equal a lot of fragility.


I've once read a HN comment that said that Aboriginees are economically useless and should work in a major city so that they can be useful. This completely ignores systematic risk posed by nuclear weapons or viruses like covid-19.

Those Aboriginees have much better survival skills than the average NYC resident and they are far away so they are more likely to survive global disasters.


My comment was about the fragility created by running organizations too lean from # of employees perspective.

I have no issue with Aborigines' life choices and don't consider them economically useless at all, so I'm not sure what that has to do with the subject.


"they are far away so they are more likely to survive global disasters"

Far away from the globe?


It's the Jack Welch algorithm


* Product is successful

* Hire lots of people because that's what successful companies do

* Product development slows way down, because you hired too fast (Mythical Man Month stuff kicks in, where adding more people to a team can slow development down, if not done carefully and thoughtfully)

* Notice that development is slow, so you hire even more people to try and speed it up

At least that's been my experience at 3 different companies.


A healthier dynamic with the same result:

- Value of a 1% improvement at $100/week revenue: $104 of developer time.

- Value of a 1% improvement at $1M/week revenue: $1,040,000 of developer time.


* Product isn't that complicated at heart, so new hires busy themselves with complicating it, to justify their paychecks

* Your easy to manage monolith is now seventy-three microservices, in case you need to serve Google levels of traffic without any further engineering. You're hopelessly understaffed to keep the monstrosity stable, so now you've got to grow your engineering corps to Google scale.


Hiring too much is a problem, but also bear in mind that having customers sort of gums up the works--you can't go changing stuff on a whim without communication, more careful assessment of impact, migration plans on customers' schedule--it's easier to be nimble when there's nothing to conserve.


100% correct.

I love the Carta philosophy on hiring - every person you need means you've failed to execute: https://carta.com/blog/how-to-hire/

Unfortunately they didn't heed their own philosophy given their layoffs with simultaneous massive fundraise.


Some of the services Patreon offers bring them dangerously close to being regulated as a lender, or even a bank. Wouldn't surprise me if some number of their staff are retained to meet any such compliance needs.


Such as?



Nothing is fully automated. At their scale even if a small fraction of their users need manual intervention that's a lot of work. I'm sure they didn't freeze their feature set 10 years ago - they have a feature backlog for sure. On top of that they appear to have an ios, android, and web app. At their scale they need an infrastructure team just to keep the plumbing working. Their administrative back end is at least as complex as their user facing front end, and probably is also constantly generating feature requests. We are easily up to 30 engineers and more support people. That means accounting, HR, managers, payroll. The beast grows.


I was at a consumer facing startup where about 70% of employees were in-house customer support representatives, so even though a product may appear simple, it can have a lot of non-programmer employees.


>take X amount of money, divide it between Y people, pay them out

What's astounding to me is they don't even do _that_! It's a many-to-one payer-payee scheme... essentially a pretty wrapper for Stripe.


> I don't really understand why Patreon needs so many employees.

Good question, I am curious about how it compares with how Gumroad operates.


Gumroad's open board meetings might give some insight into this: https://www.youtube.com/playlist?list=PL_DfN-mKCGNuswqERc6sI...


Because they took VC money and need to show fake growth.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: